Loss of Sox9 function results in defective chondrocyte differentiation of mouse embryonic stem cells...

Loss of Sox9 function results in defective chondrocyte

differentiation of mouse embryonic stem cells in vitro

GUNNAR HARGUS1,#, RALF KIST2,3,##, JAN KRAMER1,4, DANIELA GERSTEL1, ANGELA NEITZ1,GERD SCHERER2 and JÜRGEN ROHWEDEL*,1

1Dept. of Medical Molecular Biology, University of Lübeck, Lübeck, 2Institute of Human Genetics and Anthropology, University ofFreiburg, Freiburg, 3Institute of Mammalian Genetics, GSF-National Research Center for Environment and Health, Neuherberg and

4Medical Clinic I, University of Lübeck, Lübeck, Germany.

ABSTRACT The transcription factor Sox9 plays an important role during chondrogenesis. After

early conditional inactivation of Sox9 in mesenchymal limb bud cells of mice, mesenchymal

condensations as well as cartilage and bone are completely absent in the developing limbs. We

analyzed chondrogenic differentiation of Sox9-/- mouse embryonic stem cells in vitro, using two

clones with different targeted mutations. We found that the development of mature and

hypertrophic chondrocytes is completely inhibited in the absence of Sox9 confirming that Sox9

is required for the formation of cartilage. In contrast, Sox9+/- mouse embryonic stem cells showed

continous but reduced differentiation into mature chondrocytes. Interestingly, the formation of

early chondrogenic condensations expressing characteristic marker genes such as scleraxis, Sox5

and Sox6 was not inhibited in the absence of Sox9 in vitro. Thus, we propose that the earliest step

of chondrogenesis could be regulated by a non cell-autonomous function of Sox9.

KEY WORDS: chondrogenesis, in vitro differentiation, mesenchymal condensations, Sox9

Introduction

Sox9 is a member of the SOX (Sry-related high mobility groupbox) family of transcription factors that share a common 79 aminoacid DNA binding motif, known as the high-mobility group (HMG)domain, with the mammalian testis-determining factor SRY(Wegner, 1999; Bowles et al., 2000). Evidence exists that Sox9plays an important role during cartilage development (deCrombrugghe et al., 2001). During mouse embryogenesis, Sox9is expressed in all regions of cartilage formation including thesclerotomal parts of the somites that give rise to the axial skleleton,the cartilaginous elements of the limb buds which are formed bythe lateral plate mesoderm, and in neural crest-derived mesen-chymal cells of the craniofacial region (Zhao et al., 1997; Ng et al.,1997). In these structures, mesenchymal cells which differentiateinto chondroprogenitor cells and form condensations have beenshown to express Sox9 (Wright et al., 1995) as well as the celladhesion molecule N-cadherin (Oberlender and Tuan, 1994), thebasic-helix-loop-helix transcription factor scleraxis (Cserjesi etal., 1995) and the transcription factors Sox5 and Sox6 (Lefebvre

Int. J. Dev. Biol. 52: 323-332 (2008)doi: 10.1387/ijdb.072490gh

*Address correspondence to: Dr. Jürgen Rohwedel. Dept. of Medical Molecularbiology, Medical University of Lübeck, Ratzeburger Allee 160, 23538 Lübeck,Germany. Tel: +49-451-500-4095. Fax: +49-451-500-3637. e-mail: [email protected] - web address: www.molbio.uni-luebeck.de/rohwedel.htm

Present addresses: # Center for Molecular Neurobiology, University of Hamburg, D-20251 Hamburg, Germany. ## The Institute of Human Genetics, TheInternational Centre for Life, Central Parkway, Newcastle upon Tyne, NE1 3BZ, United Kingdom.

Accepted: 27th June 2007. Published online: 4th December 2007. Edited by: Christine Mummery

0214-6282/2008/$35.00© UBC PressPrinted in Spainwww.intjdevbiol.com

et al., 1998). These cells continue to express Sox9 during furtherdifferentiation, which is characterized by expression and deposi-tion of cartilage-specific matrix components such as collagen typeII and aggrecan. Sox9 is completely downregulated whenchondrocytes acquire a hypertrophic shape and start to expressCol10a1 (Zhao et al., 1997). Sox9 binds to target sites in promot-ers or enhancers of cartilage-specific genes such as Col2a1(Lefebvre et al., 1997; Zhou et al., 1998), Col11a2 (Bridgewateret al., 1998), and aggrecan (Sekiya et al., 2000), and has beenshown to activate the expression of these genes in vitro. Further-more, Col2a1 has been identified as a direct target gene of Sox9in vivo (Bell et al., 1997; Zhou et al., 1998).

Heterozygous mutations in the human SOX9 gene cause theskeletal malformation syndrome campomelic dysplasia (CD) andassociated XY sex reversal (Foster et al., 1994; Wagner et al.,1994), which establishes SOX9 as an essential factor for

THE INTERNATIONAL JOURNAL OF

DEVELOPMENTAL

BIOLOGYwww.intjdevbiol.com

Abbreviations used in this paper: EB, embryoid body; ES, embryonic stem (cell);HMG, high mobility group; HR, homologous recombination; PNA, peanutagglutinin; SOX, Sry-related high mobility group box.

324 G. Hargus et al.

skeletogenesis and for testis determination. Defects are due tohaploinsufficiency, indicating that a reduced dosage of SOX9causes the developmental abnormalities. Characteristic featuresin CD patients are defective skeletal structures such as bowingand angulation of the tibiae and femora, hypoplastic scapulae andpelvic bones, undermineralized vertebrae and craniofacial mal-formations. Possibly due to defective tracheobronchial cartilagesand narrow upper airways, most patients die in the neonatalperiod from respiratory distress (Houston et al., 1983; Mansour etal., 1995). Heterozygous Sox9-mutant mice generated by genetargeting phenocopy many of the features seen in CD patients,and also show typical abnormalities in cartilage primordia andpremature skeletal mineralization (Bi et al., 2001; Kist et al.,2002).

In mouse chimeras, Sox9-/- cells were excluded from cartilageprimordia and mature cartilage tissue throughout embryonic

development (Bi et al., 1999). Moreover, Sox9-/- teratomas didnot form any cartilage tissues (Bi et al., 1999) and both, inchimeras and teratomas, Sox9-/- cells failed to produce type IIcollagen. Furthermore, cartilage and bone was completely absentin the limbs of mice after conditional inactivation of the Sox9 genein early mesenchymal limb bud cells, and severe chondrodyspla-sia was described when Sox9 was conditionally inactivated aftermesenchymal condensations had formed (Akiyama et al., 2002).

As an alternative to homozygous Sox9 knockout mice, weanalyzed differentiation of Sox9-/- embryonic stem (ES) cells invitro to elucidate which steps of chondrocyte differentiation areaffected by loss of Sox9 function. Using the model system of EScell differentiation via embryoid bodies (EBs), we have recentlydemonstrated that the process of chondrocyte differentiation isclosely recapitulated in vitro, finally resulting in hypertrophic andcalcifying cells (Kramer et al., 2000; Hegert et al., 2002; Kramer

Fig. 1. Sox9 gene targeting strategies and inactivation by Cre-mediated recombination in ES cells. The Sox9-neoflox2 (A) and Sox9-neoflox(B) alleles, carrying loxP-flanked MC1-neo-pA cassettes (neo, open boxes) inserted into intron 1 and a third loxP site inserted into intron 2 or into the3' untranslated region, respectively, were generated by homologous recombination (HR) in ES cells. The loxP-flanked neo cassettes and either, Sox9exon 2 (A) or exon 2 and 3 (B) were removed by transient expression of Cre recombinase in vitro resulting in the Sox9-delex2 and Sox9-delex2,3 alleles,respectively. Colored boxes, Sox9 exons 1-3; triangles, loxP sites; E, EcoRI; H, HindIII. Expected restriction fragment sizes detected by thehybridisation probe 1 (A,B), or the lengths of PCR fragments amplified with primers F2 and R (B) are indicated. (A’) Southern blot analysis of theheterozygous Sox9-neoflox2 clone D4D1 shows the wildtype 7,2 kb EcoR1 and the floxed 4,8 kb EcoRI band. (B’) PCR analysis of wildtype and Sox9-neoflox ES cell clones demonstrates replacement of the wildtype allele with the mutant allele in a homozygous Sox9-neoflox clone. (C) Southern Blotanalysis of wildtype (E14.1), heterozygous D4D6 and homozygous Sox9-delex2 (D4D12-C4) and Sox9-delex2,3 (2A5-40) ES cell clones. Homologousrecombination and Cre-mediated deletion was confirmed using probe 1 showing a 6,5 kb EcoR1 and a 2,9 kb HindIII fragment in homozygous Sox9-delex2 clone D4D12-C4 and a 4,3 kb EcoRI and a 2,9 kb HindIII fragment in homozygous Sox9-delex2,3 clone 2A5-40.

EcoRI;

probe 1

7.2 -

4.8 -

D4D1

+/nf2Sox9-

wildtype

HR

Sox9-

neoflox2 (nf2)

Cre

Sox9-

del ex2

E

2 31

7.2 kb

1 2 3

E

4.8 kb

probe 1

1 3

E

6.5 kb

Sox9-

wildtype

Sox9-

neoflox

(nf)

Sox9-

del ex2,3

- 419 bp

- 247 bp

Cre

F2/R

HR

11,6 kb

E H E E E H

2 31

1 kb

3

F2 R

419 bp

247 bp

F2 R

H H

1

E H

2 3

E HEE

E

1

H HE EHE

2,9 kb

neo

probe 1

H HE

neo

H

HE H

E H

2,9 kb

7.2 kb

4,3 kb

7,2

6,5

4,3

2,9

11,6

EcoRI HindIII

E1

4.1

D4

D6

(So

x9

+/-)

D4

D12

-C4

(So

x9

-/- )

2A

5-4

0(S

ox9

-/- )

E1

4.1

D4

D6

(So

x9

+/-)

D4

D12

-C4

(So

x9

-/- )

2A

5-4

0(S

ox9

-/- )

probe 1

E H

+/+ +/nf nf/nf

2A52AE14.1

B

A' B'

C

A

Chondrogenesis of Sox9-deficient ES cells 325

et al., 2005). Here, we demonstrate that murine Sox9-/- ES cellsdifferentiate in vitro into pre-cartilage condensations which ex-press early chondrocytic marker molecules such as the transcrip-tion factors Sox5, Sox6 and scleraxis and also bind peanut-agglutinin. However, Sox9-/- cells fail to develop further and arenot able to form cartilage nodules and hypertrophic chondrocytesin vitro.

Results

Generation of Sox9+/- and Sox9-/- ES cellsBy homologous recombination with a Sox9 targeting vector,

several correctly targeted ES cell clones were identified carry-ing the modified allele Sox9-neoflox2 (Kist et al., unpublished).One of the Sox9-neoflox2 clones, D4 was used to generateSox9+/- ES cells by transient in vitro expression of Cre

D4D12-D6 were verified by Southern Blotting showing only theEcoRI 6,5 kb and the 2,9 kb HindIII fragment, respectively(Fig.1C). These subclones were used for the differentiationexperiments.

In addition, an ES cell clone with a different modified allele,Sox9-neoflox (Kist et al., 2002), was used to generate a Sox9-/-clone lacking both, exon 2 and 3 (Fig. 1B). From ES cell clone2A, carrying the Sox9-neoflox allele (Kist et al., 2002) in aheterozygous configuration, clone 2A5 was selected after cul-tivation with high concentrations of G418. PCR analysis showedthat this clone carried the Sox9-neoflox allele in a homozygousconfiguration (Fig. 1B’). Replacement of the wildtype allele onthe non-targeted chromosome by the mutant allele, is known tooccur spontaneously and has been used before to generatehomozygous ES cell clones (Mortensen et al., 1992; Lefebvreet al., 2001). After Cre expression, the subclone 2A5-40 was

Fig. 2. Loss of functional Sox9 expression in Sox9-/-

embryoid bodies. (A) RT-PCR analysis of RNA isolatedfrom wildtype (wt), heterozygous (Sox9+/-) and homozy-gous (Sox9-/-) embryoid bodies generated from ES cellclones D4D6, D4D12-C4 or 2A5-40. Oligonucleotide primersspecifically binding in exon 1 and 3 of the Sox9 gene couldnot amplify any wildtype fragment (723 bp) in both homozy-gous mutant ES cell clones. Instead, a 469 bp fragment wasdetected for the Sox9-delex2 clone D4D12-C4, while no otherfragment was detected for the Sox9-delex2,3 clone 2A5-40.(B) DNA sequencing of the 469 bp RT-PCR product revealedsplicing of exon 1 to exon 3 resulting in a frameshift muta-tion, caused by the lack of the 254 bp-long exon 2. Thisnucleotide sequence predicts an aberrant peptide whichwould code only for the first half of the HMG box (blue; DHMG-box; 42 amino acids) followed by an altered amino acidsequence and a stop codon at position 166. (C) During invitro differentiation, Sox9 protein was detected byimmunostaining in wildtype EB outgrowths (a,b) but not inSox9-/- (D4D12-C4 and 2A5-40) EB outgrowths (c,d). Repre-sentative areas are shown (wildtype and clone D4D12-C4).DIC = differential interference contrast. Scale bar, 100 µm.

recombinase. Upon complete recombination the loxP-flanked neo cassettes are removed together withexon 2, resulting in the deleted allele Sox9-delex2

(Fig. 1A). Several partially or completely recombinedES cell clones were identified by Southern blot analy-sis (data not shown). Three particular clones, D4D1,D4D6 and D4D12, were generated from the Sox9-neoflox2 cells and characterised in detail (Fig. 1 A,C).Clone D4D1 is identical to the parental clone D4 inwhich no recombination by Cre recombinase hadoccurred. The 4.8 kb EcoRI fragment in clone D4D1indicates the presence of the neo cassette in thetargeted allele (Fig. 1A’), while in clone D4D6, the 6.5kb EcoRI and the 2,9kb HindIII fragment indicate thedeleted allele (Fig. 1C). Surprisingly, in clone D4D12,the Sox9 wildtype allele on the non-targeted chromo-some had been replaced by the deleted Sox9 allele,resulting in a Sox9-/- genotype, lacking exon 2. Thisclone was subcloned because a very weak wildtypeband was detectable after prolonged exposure possi-bly due to contaminating wildtype cells (data notshown) and the resulting subclones D4D12-C4 and

N K P H V K R P M N A F M V W A Q A A R

AACAAGCCACACGTCAAGCGACCCATGAACGCCTTCATGGTGTGGGCGCAGGCTGCGCGC

R K L A D Q Y P H L H N A E L S K T L G

AGGAAGCTGGCAGACCAGTACCCGCATCTGCACAACGCGGAGCTCAGCAAGACTCTGGGC

K L W R A I S G S A D P T H H S Q N R R

AAGCTCTGGAGGGCAATCTCAGGGTCCGCCGACCCCACCCACCACTCCCAAAACCGACGT

A S W Q S -

GCAAGCTGGCAAAGTTGATCTGAAGCGAGAGGGGCGCCCTCT

N K P H V K R P M N A F M V W A Q A A R

AACAAGCCACACGTCAAGCGACCCATGAACGCCTTCATGGTGTGGGCGCAGGCTGCGCGC

R K L A D Q Y P H L H N A E L S K T L G

AGGAAGCTGGCAGACCAGTACCCGCATCTGCACAACGCGGAGCTCAGCAAGACTCTGGGC

K L W R L L N E S E K R P F V E E A E R

AAGCTCTGGAGGCTGCTGAACGAGAGCGAGAAGAGACCCTTCGTGGAGGAGGCGGAGCGG

L R V Q H K K D H P D Y K Y Q P R R R K

CTGCGCGTGCAGCACAAGAAAGACCACCCCGATTACAAGTACCAGCCCCGGCGGAGGAAG

S V K N G Q A E A E E A T E Q T H I S P

TCGGTGAAGAACGGACAAGCGGAGGCCGAAGAGGCCACGGAACAGACTCACATCTCTCCT

N A I F K A L Q A D S P H S S S G M S E

AATGCTATCTTCAAGGCGCTGCAAGCCGACTCCCCACATTCCTCCTCCGGCATGAGTGAG

V H S P G E H S G Q S Q G P P T P P T T

GTGCACTCCCCGGGCGAGCACTCTGGGCAATCTCAGGGTCCGCCGACCCCACCCACCACT

P K T D V Q A G K V D L .... P -

CCCAAAACCGACGTGCAAGCTGGCAAAGTTGATCTG.....CCTTGA

EXON 1

EXON 3

HMG-box

EXON 1

EXON 3

HMG-box�

- -/Sox9 (D4D12-C4)

101

252 507

101

165

EXON 2

wt

wt

723 bp

469 bp

�-tubulin

Sox9

--/

Sox9

(D4D

6)

--/

Sox9

-+/

Sox9

(D4D

12-C

4)

(2A

5-4

0)

Sox9DIC

- -/Sox9

wt

- -

a b

c d

B

CA


selected which only carried the Sox9-delex2,3 alleles as verifiedby Southern Blotting showing 4,3 kb EcoRI and 2,9 kb HindIIIfragments, respectively (Fig. 1C).

To investigate the Sox9 transcripts generated from theSox9-delex2 allele, we performed RT-PCR with RNA isolatedfrom ES cell clones D4D6 and D4D12-C4 and sequence-specific primers located in exon 1 and exon 3. We found that thefragment amplified from the transcripts of the deleted allele(469 bp) was 254 bp shorter than the wildtype fragment (723 bp)and no Sox9 wildtype transcripts could be detected in cloneD4D12-C4 (Fig. 2A). DNA sequencing of the mutant 469 bpfragment showed that exon 2 was completely deleted and exon1 was spliced to exon 3 resulting in a frameshift mutation. If aprotein was to be translated from the mutant transcript of theSox9-delex2 allele, it would consist of 165 instead of 509 aminoacids, carrying a truncated HMG domain followed by 21 out-of-frame amino acids encoded by exon 3 (Fig. 2B). RT-PCR withRNA isolated from ES cell clone 2A5-40 carrying the Sox9-delex2,3 allele did not result in any amplification product (Fig.2A). By immunostaining we were not able to detect any Sox9protein in the knock-out clones D4D12-C4 (Fig. 2C) and 2A5-40

(data not shown) during differentiation.

Sox9-/- cells are unable to develop into mature chondrocytesin vitro

ES cells differentiate in vitro into highly organized cartilagenodules (Kramer et al., 2000). To test for the chondrogenic in vitrodifferentiation capacity in the absence of Sox9, wildtype, Sox9+/-and Sox9-/- ES cells were analyzed by counting type II collagen-positive cartilage nodules (Fig. 3 A,B) in EB outgrowths. We foundthat cells of the Sox9-/- clones D4D12-C4 and D4D12-D6 and ofthe different Sox9-/- clone 2A5-40 did not differentiate into thesehighly organized cartilage structures during EB cultivation up to5+31 d (Fig. 3C). In contrast, in Sox9+/- EB outgrowths, nodulesappeared but the number was reduced in comparison to thewildtype control (Fig. 3C). The first nodules were detected inwildtype EBs at 5+13 d, their number increased up to 5+17 d anddecreased thereafter. Similarly, in Sox9+/- EB outgrowths, thefirst nodules formed at 5+15 d, the maximum number was de-tected at 5+19 d and decreased later. The mean values for thenumber of type II collagen–positive nodules in Sox9+/- EB out-growths never reached the wildtype levels, although the differ-

Fig. 3 (Left). Sox9-/- ES cells are not able to differentiate into mature chondrocytes forming cartilage nodules. The number of cartilage nodules(A,B,D,E) was determined in embryoid body (EB) outgrowths of wildtype (wt), Sox9+/- (D4D6) and Sox9-/- (D4D12-C4, D4D12-D6 and 2A5-40) ES cellsafter immunostaining for type II collagen (A-C) and after Alcian blue staining (D-F). Whereas in wildtype EBs, the first nodules were found 13 daysafter plating (5+13 d), and two days later in Sox9+/- EBs (5+15 d), no such structures could be detected in Sox9-/- EBs (C,F). In Sox9+/- EBs, the numberof nodules appeared to be reduced but the differences were not statistically significant. Mean values ± SEM from at least three independentexperiments are shown. Because we found less than 1 nodule per EB in the wildtype EBs we analyzed approximately 200 EBs per day. Significantdifferences between wildtype and Sox9-/- EBs: *: p < 0.05; **: p < 0.01. Scale bars, 100 µm.

Fig. 4 (Right). Wildtype and Sox9+/- embryoid bodies (EBs) form nodules consisting of hypertrophic chondrocytes. Wildtype (A,C,E,G) andSox9+/- (B,D,F,H) ES cells differentiated into Col10a1- (A,B) and type II collagen (C,D)-positive cells. These cells were embedded in abundant extracellularmatrix and formed cartilaginous nodules (E,F). Merging of the pictures demonstrates that type II collagen is evenly distributed between the Col10a1-expressing cells derived from wildtype and Sox9+/- ES cells (G,H). Sox9-/- ES cells do not form cartilage nodules. EBs were analyzed at 5+28 d by in situhybridization for Col10a1 mRNA combined with immunostaining for type II collagen. DIC, differential interference contrast. Scale bar, 100 µm.

Time of differentiation (d)

5+ 5 10 15 20 25 30 35

0

0.1

0.2

0.3

0.4

0.5


5+ 5 10 15 20 25 30 35

0

0.1

0.2

0.3

0.4

0.5wildtype

Sox9 -/- (D4D12-C4, D4D12-D6 and 2A5-40)Sox9 +/-

** ** ** ***

** ** ***

Co

llag

en

II-

po

sitiv

en

od

ule

sp

er

EB

Alc

ian

blu

e-

sta

ine

dn

od

ule

sp

er

EB

wt

wt

wt

wt

wildtype

Sox9 -/- (D4D12-C4, D4D12-D6 and 2A5-40)Sox9 +/-

B

C

D

E

A

F

wildtype Sox9+/-

Col10a1

type II col

DIC

merge

A B

C D

E F

G H


ences were not statistically significant. These results were con-firmed by Alcian blue staining, which showed that deep bluestained nodules (Fig. 3D,E) were completely absent in Sox9-/- EBoutgrowths up to 5+33 d, and that their number was reduced inSox9+/- EBs (Fig. 3F). Alcian blue specifically stains acidicproteoglycans found in the extracellular matrix of cartilage tissue.The maximum number of Alcian blue-stained nodules was ob-served at later stages as compared to type II collagen-positivenodules, in both wildtype and Sox9+/- EBs. This is probably dueto the fact that the Alcian blue-stained proteoglycans and type IIcollagen are subsequently expressed during terminal chondro-genic differentiation in EBs. In summary, these data demonstratethat Sox9 is required for the formation of highly organized carti-lage nodules in vitro.

To characterize the terminal stage of chondrocyte differentia-tion in more detail, the expression of Col10a1 was analyzed, amarker for hypertrophic chondrocytes appearing at the latestdifferentiation steps during endochondral differentiation in vivo. Insitu hybridization for Col10a1 mRNA (Fig. 4 A,B) combined withimmunostaining for type II collagen (Fig. 4 C,D) demonstrated thathypertrophic chondrocytes co-expressing Col10a1 and type IIcollagen were present in wildtype and Sox9+/- EBs but not inSox9-/- EBs (not shown). These results show that chondrogeniccells in Sox9+/- ES cell-derived nodules become hypertrophic.Hypertrophic cells of wildtype and Sox9+/- EBs did not show anyobvious differences regarding their morphology or expressionand distribution of type II collagen (Fig. 4 C-F). However, in situhybridization showed that expression of Col10a1 was alwaysstronger in Sox9+/- EBs and the number of Col10a1-expressingcells seemed to be increased compared to the wildtype (Fig. 4A,B).

Sox9-/- ES cells are able to form pre-cartilage condensationsin vitro

To see whether already early steps of chondrogenic differen-tiation were affected after complete loss of Sox9 function in vitro,we tested the EBs for the presence of pre-cartilage condensa-tions. To this end we analyzed wildtype and Sox9-/- (clonesD4D12-C4, and 2A5-40) EB outgrowths for expression of Sox5(Fig. 5 A-C) and Sox6 (Fig. 5 J-L) and found that both markers arecoexpressed with type II collagen in pre-cartilage condensationsof wildtype and Sox9-/- EB outgrowths (Fig. 5). Control slideshybridized with the respective sense probes and the secondaryantibody alone were negative (data not shown). Condensations inboth, wildtype and Sox9-/- (D4D12-C4, and 2A5-40) EB out-growths, displayed intensive binding of PNA (Fig. 5 S-U), demon-strating that pre-cartilage condensations are formed in the ab-sence of Sox9. Identical results were obtained for the Sox9-/-clone D4D12-D6 (data not shown).

To analyze condensation formation on a quantitative level, weanalyzed wildtype, Sox9+/- and Sox9-/- (D4D12-C4) EBs for theformation of N-cadherin- and scleraxis-positive cell condensa-tions. We found scleraxis- (Fig. 6 A-C) and N-cadherin- (Fig. 6 D-F) positive cell condensations in EB outgrowths of wildtype,Sox9+/- and Sox9-/- ES cells during cultivation. There were noobvious morphological differences between these pre-cartilagestructures among the three genotypes and we did not find anysignificant quantitative differences in the number of scleraxis-positive condensations between wildtype, Sox9+/- and Sox9-/-

Fig. 5. Pre-cartilage condensations show expression of Sox5 and

Sox6 and bind peanut agglutinin (PNA) in the absence of Sox9. In situhybridization for Sox5 (A-C) and Sox6 mRNA (J-L) combined withimmunostaining for type II collagen (D-F, M-O) and DIC microscopy (G-

I, P-R) showed that mesenchymal cells in wildtype (wt; A,D,G,J,M,P) andSox9-/- EBs derived from the two different Sox9-/-ES cell clones D4D12-C4 (B,E,H,K,N,Q) and 2A5-40 (C,F,I,L,O,R) form condensations whichexpress Sox5 and Sox6. Such condensations also bind fluorescein-labeled PNA in wildtype (S,V) and Sox9-/- EBs derived from clonesD4D12-C4 (T,W) and 2A5-40 (U,X). Representative areas from EBoutgrowths are shown. Scale bar, 100 µm.

EBs (Fig. 6 J). The same results were obtained when we studiedexpression of scleraxis and N-CAM, which are also characteris-tically co-expressed by cells forming pre-cartilage condensations(data not shown).

Together, these results indicate that Sox9 inactivation does

Sox5

wt

type II col

DIC

-/- (D4D12-C4) -/- (2A5-40)

Sox6

DIC

DIC

PNA

type II col

A B C

D E F

G H I

J K L

M N O

P Q R

S T U

V W X


lated in the absence of Sox9 (Fig. 7 B). Thus, Sox9 seems to berequired for an upregulated expression of Col2a1 but otherfactors may compensate for a basal expression at least in vitro.

Discussion

Sox9 is required for ES cell differentiation into cartilagenodules in vitro

Our results demonstrate that in vitro differentiation of Sox9-/-ES cells into chondrocytes is disrupted at a stage characterisedby the appearance of typical round-shaped chondrocytes orga-nized in distinct nodules and expressing a high level of type IIcollagen. Sox9-/- cells fail to form these cartilage nodules. In linewith this, it has been demonstrated that Sox9-deficient cells couldnot differentiate into mature chondrocytes in mouse Sox9-/-chimeras and teratomas (Bi et al., 1999). Furthermore, condi-tional null mutant mice lacking Sox9 in mesenchymal cells of limbbuds were not able to form cartilage and bone in limbs (Akiyamaet al., 2002). Since we found in previous studies that the formationof cartilage nodules in differentiating EBs can be induced afterapplication of growth factors of the transforming growth factor β-family (Kramer et al., 2000) we applied BMP-2, TGF-β1 and TGF-β3 to differentiating Sox9-/- EBs. However, we did not detect anycartilage nodule in these EB outgrowths (data not shown) indicat-ing that loss of Sox9 function cannot be rescued by thesechondrogenic factors in vitro. In Sox9+/- EB outgrowths, cartilagenodules were present, but the number of nodules appeared to bereduced compared to the wildtype, indicating that Sox9 genedosage is important for proper cartilage differentiation. Thisagrees with the observation that cartilage structures are defectiveand hypoplastic but not completely absent in CD patients andSox9+/- mice (Houston et al., 1983; Bi et al., 2001). Takentogether, this demonstrates that Sox9 plays an essential roleduring chondrogenesis both in vivo and in vitro.

It has been proposed that Sox9 plays an inhibitory role for theswitch from prehypertrophic to hypertrophic chondrocytes, be-cause Sox9 is completely switched off in hypertrophic chondro-cytes in vivo (Zhao et al., 1997; Bi et al., 2001). Furthermore, inprehypertrophic chondrocytes of mice lacking the receptor for theparathyroid hormone-related peptide, Sox9 phosphorylation isabolished and these mice show accelerated differentiation ofhypertrophic chondrocytes (Lanske et al., 1996; Huang et al.,2001). Our data suggest that Col10a1 expression and the number

not affect the formation of early pre-cartilage condensationsduring ES cell differentiation in vitro.

Col2a1 expression is downregulated but not completelyabolished during in vitro ES cell differentiation in the ab-sence of Sox9

Because Sox9 is an activator of the Col2a1 gene both in vivoand in vitro, we performed conventional and quantitative RT-PCRof Sox9-deficient EBs. Using primer which amplify two splicevariants of Col2a1, a juvenile and an adult form (Metsäranta et al.,1991), we found that both were expressed in Sox9-/- EBs ofclones D4D12-C4 and 2A5-40 (Fig. 7A). This was confirmed byimmunostaining of EB cultures for collagen type II. Both, inwildtype and Sox9-/- EBs, type II collagen-fibers could be de-tected at late differentiation stages, scattered throughout the EBoutgrowths (Fig. 7 C,D). These fibers were also found in Sox9+/-EB outgrowths (not shown). To determine the level of Col2a1expression in wildtype and Sox9-/- cells, EBs of clone 2A5-40 andD4D12-C4 were analyzed by real-time RT-PCR for expression ofCol2a1 using primers which do not discriminate between bothsplice variants. Col2a1 was still expressed but clearly downregu-

Fig. 6. Loss of Sox9 function does not affect the formation of pre-

cartilage condensations in vitro. In situ hybridization using an anti-sense probe detecting scleraxis mRNA (A,B,C) combined withimmunostaining for N-Cadherin (D,E,F) and DIC microscopy (G,H,I)

showed that mesenchymal cells in wildtype (wt; A,D,G), Sox9+/- D4D6(B,E,H) and Sox9-/- D4D12-C4 (C,F,I) EBs form condensations (G,H,I)

which express scleraxis (A,B,C) and N-Cadherin (D,E,F). Representativeareas from EB outgrowths are shown. DIC, differential interferencecontrast. Scale bar, 100 µm. Scleraxis-positive pre-cartilage condensa-tions were found in EB outgrowths of the three ES cell clones during invitro differentiation from 5 days (5+5 d) up to 31 days (5+31 d) after EBplating (J). Data are shown for wildtype, the heterozygous clone D4D6and the homozygous clone D4D12-C4. The number of scleraxis-positivecondensations did not differ significantly between EBs of the threegenotypes. Mean values from at least three independent experimentsare shown. Approximately 50 EBs were analyzed per time point.

J


5+ 5 10 15 20 25 30 35

0

2

4

6

8

wt

Sox9-/-

Sox9+/-

scleraxis N-cadherin

wt

-/-

+/-

DIC

Scle

raxis

-po

sitiv

eco

nd

en

sa

tio

ns

pe

rE

B

A

B

D G

E H

C F I


of Col10a1- expressing cells increase in Sox9+/- nodules. Theseresults indicate that a reduced level of Sox9 may promote theformation of hypertrophic chondrocytes. In line with this, Sox9+/-mice showed an enlarged zone of chondrocyte hypertrophy in thegrowth plates of the long bones (Bi et al., 2001).

Formation of pre-cartilage condensations is not affected inSox9-/- EBs

We found that during ES cell differentiation in vitro, Sox9-deficient cells form early pre-cartilage condensations which ex-press Sox5, Sox6, scleraxis and N-cadherin, markers character-istic for such pre-cartilage condensations (Oberlender and Tuan,1994; Lefebvre et al., 1998; Brown et al., 1999). Furthermore, thecondensations strongly bind the lectin PNA (DeLise et al., 2000).These results were not only obtained with both subclones from theSox9-/- clone D4D12, lacking exon 2, but also with the indepen-dent Sox9-/- clone 2A5-40, lacking exons 2 plus 3.

In mouse chimeras, Sox9-/- cells were located adjacent tocondensing wildtype mesenchymal cells in 11.5 and 12.5 d p.c.embryos and did not take part in the formation of mesenchymalcondensations (Bi et al., 1999). Moreover, conditional null mutantmice, in which Sox9 was inactivated in mesenchymal cells of limbbuds before these cells condense, were no longer able to formmesenchymal condensations (Akiyama et al., 2002). These re-sults suggested that condensation formation is a cell-autono-mous process under the control of Sox9. On the other hand,zebrafish with homozygous null mutations in the sox9a gene, anortholog of the mammalian Sox9 gene, are able to form mesen-chymal condensations in the first two pharyngeal arches (Yan etal., 2002). We have shown in the present study, that a homoge-neous population of Sox9-deficient cells is able to form pre-cartilage condensations and express early molecular markers invitro. Thus, our data indicate that the function of Sox9 duringcondensation formation is rather not cell-autonomous.

In vivo, development is not only temporally controlled as inEBs, but is also spatially controlled by a combination of distinctsignaling molecules present at defined concentrations. Becausemorphogenetic development is not possible within EBs, suchspatially controlled signals might be lacking, resulting in conden-sation of mesenchymal cells in Sox9-/- EBs. In line with this, avariability of morphogenetic signals around pre-cartilage con-densations in EB outgrowths might influence their cellular fate. Infact, we found that in wildtype EBs only some of the condensa-tions develop into cartilage nodules, and the mean number ofscleraxis-positive condensations does not decrease significantlyduring culture. Thus, the in vitro system offers the possibilty toanalyze the potency of cell-autonomous differentiation effects.The formation of specifically shaped mesenchymal condensa-tions in vivo may be regulated by antagonistic factors produced byectodermal or non-condensing mesenchymal cells located closeto the place of the condensations (Zanetti and Solursh, 1986). Forexample, in limb buds, the size and shape of mesenchymalcondensations is controlled by inhibitory factors produced byectodermal cells such as FGF2 and FGF8 (Moftah et al., 2002).Sox9 could be an antagonist of such condensation-inhibitingfactors, which are expressed in vivo by the overlaying ectoderm.The absence of Sox9 would then result in the complete loss ofcondensations in vivo. In contrast, in the EB in vitro differentiationsystem, the release of such condensation-inhibiting factors from

adjacent tissue can not occur and therefore condensations canform even in the absence of Sox9. Similarly, it has been found thatscleraxis null mutant embryos fail to form mesoderm, whereas inEBs of scleraxis-/- ES cells, mesodermal markers were ex-pressed at a similar level as in wildtype EBs (Brown et al., 1999)indicating that the role of scleraxis during mesoderm formation isnot cell-autonomous but depends on the environment. Anotherexample are mice lacking a transcription factor, the serum re-sponse factor (Srf). These Srf-/- mice stop developing at the onsetof gastrulation and do not form mesoderm (Arsenian et al., 1998).However, Srf-/- ES cells differentiated in vitro into mesodermalcell types although this process was impaired in vivo (Weinhold etal., 2000) suggesting that the function of Srf to promote meso-derm formation is non-cell-autonomous. Our data indicate that anon cell-autonomous function may also apply to Sox9 regardingthe formation of pre-cartilage condensations.

Loss of Sox9 affects the level of but does not completelyabolish Col2a1 expression in vitro

Expression studies in vivo suggested that Col2a1 is a target forSox9, because Sox9 and Col2a1 are coexpressed in cartilageprimordia throughout the developing skeleton and in other devel-oping cartilage structures during embryogenesis (Zhao et al.,1997; Ng et al., 1997). It has also been shown that Sox9 binds to

Fig. 7. Expression of Col2a1 is clearly downregulated, but not com-

pletely abolished, during in vitro differentiation of Sox9-deficient ES

cells. Both the adult (225 bp) and the juvenile (432 bp) splice variants ofCol2a1 were expressed in the absence of Sox9 as analyzed by conven-tional RT-PCR in 5+22 d EBs derived from wildtype (wt) and Sox9-deficientclones D4D12-C4 and 2A5-40 (A). However, the level of gene expressionof Col2a1 as analyzed by real time RT-PCR during differentiation ofwildtype (wt) and Sox9-/- ES cells (clones D4D12-C4 and 2A5-40) wasclearly reduced in Sox9-/- EBs compared to wildtype EBs (B). Data areshown for clone D4D12-C4. RNA isolated from undifferentiated ES cells(0d) and EBs at 5d, 5+7 d, 5+14 d and 5+24 was analyzed. Immunostainingsshowed that collagen II fibres were scattered throughout EBs derivedfrom wildtype (C) and Sox9-/- ES cells (D) at 5+28 d. Collagen type II fibreformation is independent of cartilage nodule formation which is missing inthe absence of Sox9. Scale bar, 100 µm.

0d 5d plus 7 plus 14 plus 24

0

10

20

30

40

50

60

ng

mR

NA

/ml


wt

Sox9-/- (D4D12-C4)

wt

Sox

9-/-

(D4D

12-C

4)

Sox

9-/-

(2A5-

40)

423

225

bp

wt

Sox9 -/-

C

D

A

B


specific sequence elements of the Col2a1 enhancer and directschondrocyte-specific Col2a1 expression, both in transient trans-fection experiments and in transgenic mice (Lefebvre et al., 1996;Bell et al., 1997; Lefebvre et al., 1997; Zhou et al., 1998). In mousechimeras, Sox9-/- cells did not express Col2a1, and in Sox9-/-teratomas, type II collagen was not detectable in any cell type (Biet al., 1999). In contrast, coexpression studies of Sox9 andCol2a1 in developing wild type mouse embryos revealed thatCol2a1 was expressed in several nonskeletal tissues which arenegative for Sox9 (Ng et al., 1997), indicating that differentiatingcells are able to express Col2a1 in the absence of Sox9. Further-more, type II collagen expression in cultured human articularchondrocytes does not correlate with the level of Sox9 expression(Aigner et al., 2003). We found that Sox9-deficiency did not resultin complete abolishment but obvious downregulation of Col2a1expression in differentiating EBs, as shown by real-time RT-PCRand by detection of type II collagen fibres in Sox9-/- EB out-growths. One possible explanation for this unexpected resultcould be that we generated a partially functional Sox9 protein byour gene targeting strategy. This can be ruled out, as we did notdetect any Sox9 protein in cultured Sox9-/- EBs. Furthermore, ithas been shown recently that no functional Sox9 protein could bedetected after conditional inactivation of Sox9 in the lung using thesame targeting strategy resulting in Sox9-delex2,3 alleles (Perl etal., 2005). Moreover, even if a truncated protein was still producedin the knock-out lines this mutant Sox9 protein would lack afunctional DNA binding domain and the C-terminal transactivationdomain (Südbeck et al., 1996) and would thus not be able tofunction as a transcription factor.

Another explanation would be compensation of Sox9 functionin vitro by a protein with an overlapping function. The transcriptionfactors Sox5 and Sox6 would be candidates. Both were ex-pressed in Sox9-deficient condensations in vitro and might up-regulate Col2a1 expression at least to moderate expressionlevels. Such a compensatory mechanism may depend on theprior formation of mesenchymal condensations. This agrees withthe observation that in conditional null mutant mice, expression ofCol2a1 as well as Sox5 and Sox6 was inhibited when Sox9 wasinactivated in mesenchymal cells before condensations had beenformed (Akiyama et al., 2002).

In conclusion, this in vitro study unravels a mechanistic insightinto the function of Sox9 during chondrogenic differentiation. Wefound that in contrast to the terminal step of differentiationcharacterized by the formation of cartilage nodules, the early stepof pre-chondrogenic differentiation, the formation of pre-cartilagecondensations, remained almost unaffected after loss of Sox9function in vitro. In contrast, a block of this early differentiationstep has previously been demonstrated in vivo. This indicates thatthe function of Sox9 during this process is not cell-autonomous.Thus, the in vitro differentiation of embryonic stem cells is a usefulapproach to bring new important insights into complex develop-mental processes.

Materials and Methods

Generation of Sox9+/- and Sox9-/- ES cellsGene targeting of Sox9 was achieved by electroporation of E14.1 ES

cells (Kuhn et al., 1991) with a targeting vector containing a loxP-flankedMC1-neo-pA cassette inserted into intron 1, a third loxP site inserted intointron 2, and a PGK-tk-pA negative selection cassette (Kist et al.,

unpublished data) generating the Sox9-neoflox2 allele. The targetingvector is thus largely identical to the targeting vector described previouslywhich resulted in the Sox9-neoflox allele (Kist et al., 2002), except that thethird loxP site is placed within intron 2 instead of downstream of exon 3.Cultivation, electroporation and drug selection of E14.1 ES cells wasaccording to standard procedures (Matise et al., 2000). Correctly targetedES cell clones were identified and confirmed by Southern blot analysisusing specific hybridisation probes (Kist et al., unpublished). In order togenerate Sox9+/- ES cell clones by transient in vitro expression of Crerecombinase, correctly targeted Sox9 ES cell clones were electroporatedwith pCre-Pac expression vector and selected with puromycin as de-scribed (Taniguchi et al., 1998). Fortuitously, a Sox9-/- ES cell clone,termed D4D12, was obtained in which the wildtype allele on the non-targeted chromosome had been replaced by the mutant allele lackingexon 2. The underlaying mechanisms of such chromosome-specific lossof heterozygosity has been discussed elsewhere (Lefebvre et al., 2001).

A different Sox9-/- ES cell line was generated from a clone, termed 2A,carrying the Sox9-neoflox allele in a heterozygous configuration (Kist etal., 2002). After drug selection with high concentrations of G418 of 10 and12 mg/ml for 38 days, surviving clones were picked, genomic DNA wasisolated and screened by PCR for homozygosity of the Sox9-neofloxallele. A clone, termed 2A5, was obtained which carried two Sox9-neofloxalleles. This clone was transfected with the pCre-Pac expression vectorand cells were selected with puromycin as described above.

For subcloning, Sox9-/- ES cell clones D4D12 and 2A5 were plated atlow density and new subclones, termed D4D12-C4, D4D12-D6 and 2A5-40, were isolated and confirmed by Southern Blotting.

Cell culture and differentiation of EBsDifferentiation of chondrogenic cells in vitro was studied during differ-

entiation of the Sox9+/- ES cell clone D4D6 and the Sox9-/- ES cell clonesD4D12-C4 and D4D12-D6, lacking exon 2, and 2A5-40, lacking exon 2and exon 3, in comparison to the wildtype ES cell line E14.1. ES cells weregrown on a feeder layer of mitomycin C-inactivated mouse embryonicfibroblasts in cultivation medium consisting of DME (INVITROGEN,Karlsruhe, FRG) supplemented with 15% FCS (INVITROGEN, Karlsruhe,FRG), non-essential amino acids (INVITROGEN, Karlsruhe, FRG, stocksolution diluted 1:100), 2 mM L-glutamine (INVITROGEN, Karlsruhe,FRG) and 5x10-5 M β-mercaptoethanol (SERVA, Heidelberg, FRG), asdescribed previously for line D3 (Kramer et al., 2000). For differentiation,aliquots of 20 µl differentiation medium (containing 20% FCS instead of15%) containing 800 cells were cultivated in «hanging drops» for 2 daysand, after transfer on bacteriological petri dishes, in suspension for anadditional 3 days (Kramer et al., 2000). The 5 day («5 d») old EBs wereplated separately onto gelatin (0.1%)-coated 24 well microwell plates formorphological analysis, or 15 EBs were plated onto a 6 cm tissue cultureplate for Alcian blue staining and RT-PCR, or 10 EBs onto 2 well (21,3 x20 mm) Lab-Tek chamber slides (NUNC, Wiesbaden, FRG) forimmunostaining, in situ hybridization and test for PNA binding. Alcian bluestainings were performed as described previously (Kramer et al., 2000).We performed at least three differentiation experiments per cell line andanalyzed approximately 200 EBs per time point by Alcian blue staining.Data analysis was performed using the Sigma Plot 5.0 software (JANDEL,Corte Madeira, USA). The Student’s t-test was used for statistical analy-sis.

Quantitatitve measurement of Col2a1 gene expression by real-time-RT-PCR analysis

Samples of ten EBs of different developmental stages up to 26 daysafter plating (5+26 d) were collected, washed two times with PBS, andtotal RNA was isolated and reverse transcribed as described (Hegert etal., 2002). Aliquots of 1 µl from the RT reactions were mixed with 10pmolprimer specific for Col2a1 (sense: 5'-TTTCCTCCGTCTACTGTCCACTG-3'-; antisense: 5'-TGTATGTGAACCTGCTGTTGC≠C-3'; product size:161bp) and real-time PCR was carried out with the iQ SYBR Green


Supermix (BIORAD, Munich, FRG) using an iCyler iQ thermal cycler(BIORAD, Munich, FRG) according to the manufacturer‘s instructions.The thermal cycling conditions were 95 °C for 2 minutes followed by 40cycles of 95°C for 40 seconds, 58 °C for 40 seconds and 72 °C for 40seconds. To confirm the specificity of the amplified products, meltingcurves were performed at the end of the amplification by coolingsamples to 58 °C for 1 minute and then increasing temperature to 95°C at 0.05 °C/second with continuous fluorescence measurement.Each sample was tested in duplicate. For generation of standardcurves, the PCR product was cloned into the vector pCR-TOPO(INVITROGEN, Karlsruhe, FRG). Plasmid DNA was isolated usingQIAGEN-tip 100 anion-exchange columns (QIAGEN) and serially di-luted in double-distilled water. Threshold cycles are adjusted to attainthe highest possible correlation coefficient value for the standard curveprovided by the manufacturer’s software. According to their respectivecycle numbers the concentrations of unknown samples were deducedfrom the standard curve.

Conventional RT-PCR analysisRT-PCR reactions were carried out with sequence-specific primers

as described previously (Kramer et al., 2000; Hegert et al., 2002). Tostudy expression of Col2a1 and Sox9, respectively, the followingprimers were used (oligonucleotide sequences are given in brackets inthe order antisense-, sense-primer followed by the annealing tempera-ture used for PCR, length of the amplified fragment and a reference):Col2a1 5' - AGGGGTACCAGGTTCTCCATC - 3',

5' - CTGCTCATCGCCGCGGTCCTA - 3'; 60°C; 432 bp (splicevariant A) and 225 bp (splice variant B); (Metsäranta et al., 1991);Sox9 5' – TGGGTGGCAAGTATTGGTCAAACTCA –3',

5' – TGAAGAAGGAGAGCGAGGAAGATAA 3'; 57°C; 723bp; (Lefebvre et al., 1998). Electrophoretic separation of PCR productswas carried out on 2% agarose gels.

Cloning of PCR fragments for sequence analysisReverse Transcription was performed as described above. cDNA

was amplified by PCR using Vent-DNA-polymerase (NEW ENGLANDBIOLABS, Frankfurt, FRG). 723 bp and 469 bp long Sox9 cDNAfragments amplified from RNA isolated from wildtype and Sox9-/-(clone D4D12-C4) cells, respectively, were excised from a 2% agarosegel after electrophoretic separation, and purified using the QIAquickGel Extraction Kit (QIAGEN, Hilden, FRG). These fragments werecloned blunt-ended into the plasmid vector pCR®-Blunt using the ZeroBlunt™ PCR Cloning Kit according to the manufacturer‘s protocol(INVITROGEN, Karlsruhe, FRG). Clones carrying inserts of the ex-pected length were selected after restriction enzyme digestion andtheir nucleotide sequence was verified by sequencing (MWG-Biotech,Ebersberg, FRG).

Fluorescence in situ hybridisation coupled with immunostainingThe combination of fluorescence in situ hybridization and

immunostaining as well as cloning of the scleraxis and Col10a1 cDNAsused to generate RNA probes by in vitro transcription have beendescribed previously (Kramer et al., 2000). The probes used to detectSox5 and Sox6 have been described elsewhere (Lefebvre et al., 1998).Ten EBs were plated per chamber slide and analyzed at differentdevelopmental stages. The monoclonal antibody II-II6B3 (Develop-mental Studies Hybridoma Bank, University of Iowa, USA) against typeII collagen or the monoclonal anti-A-CAM antibody GC-4 (SIGMA,Taufkirchen, FRG) to detect N-Cadherin was applied in a 1:20 or 1:40dilution, respectively. For immunostaining against Sox9 we used apolyclonal antibody from Santa Cruz (Heidelberg, FRG) diluted 1:200.Secondary antibodies were 1:100 diluted FITC-conjugated sheep F(ab)fragments against digoxigenin (BOEHRINGER, Mannheim, FRG) andCy3- or FITC-conjugated goat anti-mouse or anti-rabbit secondaryantibodies (DIANOVA, Hamburg, FRG) diluted 1:800 in PBS. Control

slides were hybridized with the respective sense probes and thesecondary antibody alone.

PNA bindingTo test for binding of Peanut agglutinin (PNA) to pre-cartilage

condensations, EBs plated onto chamber slides were washed threetimes with PBS, fixed in 3.7% formaldehyde in PBS for 30 minutes atroom temperature and washed again three times with PBS. FITC-labeled PNA (BIOMEDA, Foster City, USA) was applied at a concen-tration of 0.1 mg/ml and incubated for 45 minutes at room temperature.After washing four times in PBS, specimen were embedded inVectashield mounting medium (VECTOR, Burlinggame, USA).

Acquisition and processing of imagesSlides were analyzed with the fluorescence microscope AXIOSKOP

(ZEISS, Oberkochen, FRG) equipped with a 3 CCD color video camera(SONY, Cologne, Germany) using the acquisition software AXIOVISION(ZEISS, Oberkochen, FRG). Figures were assembled using the CORELDRAW software (COREL Corp., Ottawa, Canada).

AcknowledgementsThe skilful technical assistance of A. Eirich and M. Dose and is

gratefully acknowledged. The II-II 6B3 monoclonal antibody developed byT. F. Linsenmayer was obtained from the Developmental Studies Hybri-doma Bank developed under the auspices of the NICHD and maintainedby the University of Iowa, Dept. of Biological Sciences, Iowa City,IA52242. The work was supported by grants from the DeutscheForschungsgemeinschaft to J.R. (Ro 2108/1-1 and 1-2) and to G.S. (Sche194/11-3) and by funding from Intermed Service GmbH, Geesthacht. Theprobes used for Sox5 and Sox6 mRNA in situ hybridization were kindlyprovided by V. Lefebvre (Cleveland Clinic Foundation, Cleveland, OH,U.S.A.).

References

AIGNER, T., GEBHARD, P.M., SCHMID, E., BAU, B., HARLEY, V. and POSCHL,E. (2003). SOX9 expression does not correlate with type II collagen expressionin adult articular chondrocytes. Matrix Biol. 22: 363-372.

AKIYAMA, H., CHABOISSIER, M.C., MARTIN, J.F., SCHEDL, A. and DECROMBRUGGHE, B. (2002). The transcription factor Sox9 has essential rolesin successive steps of the chondrocyte differentiation pathway and is requiredfor expression of Sox5 and Sox6. Genes Dev. 16: 2813-2828.

ARSENIAN, S., WEINHOLD, B., OELGESCHLAGER, M., RUTHER, U. andNORDHEIM, A. (1998). Serum response factor is essential for mesodermformation during mouse embryogenesis. EMBO J. 17: 6289-6299.

BELL, D.M., LEUNG, K.K., WHEATLEY, S.C., NG, L.J., ZHOU, S., LING, K.W.,SHAM, M.H., KOOPMAN, P., TAM, P.P. and CHEAH, K.S. (1997). SOX9directly regulates the type-II collagen gene. Nat. Genet. 16: 174-178.

BI, W., DENG, J.M., ZHANG, Z., BEHRINGER, R.R. and DE CROMBRUGGHE, B.(1999). Sox9 is required for cartilage formation. Nat. Genet. 22: 85-89.

BI, W., HUANG, W., WHITWORTH, D.J., DENG, J.M., ZHANG, Z., BEHRINGER,R.R. and DE CROMBRUGGHE, B. (2001). Haploinsufficiency of Sox9 resultsin defective cartilage primordia and premature skeletal mineralization. Proc.Natl. Acad. Sci. U. S. A 98: 6698-6703.

BOWLES, J., SCHEPERS, G. and KOOPMAN, P. (2000). Phylogeny of the SOXfamily of developmental transcription factors based on sequence and structuralindicators. Dev. Biol. 227: 239-255.

BRIDGEWATER, L.C., LEFEBVRE, V. and DE CROMBRUGGHE, B. (1998).Chondrocyte-specific enhancer elements in the Col11a2 gene resemble theCol2a1 tissue-specific enhancer. J. Biol. Chem. 273: 14998-15006.

BROWN, D., WAGNER, D., LI, X., RICHARDSON, J.A. and OLSON, E.N. (1999).Dual role of the basic helix-loop-helix transcription factor scleraxis in mesodermformation and chondrogenesis during mouse embryogenesis. Development126: 4317-4329.

CSERJESI, P., BROWN, D., LIGON, K.L., LYONS, G.E., COPELAND, N.G.,


GILBERT, D.J., JENKINS, N.A. and OLSON, E.N. (1995). Scleraxis: a basichelix-loop-helix protein that prefigures skeletal formation during mouse embryo-genesis. Development 121: 1099-1110.

DE CROMBRUGGHE, B., LEFEBVRE, V. and NAKASHIMA, K. (2001). Regulatorymechanisms in the pathways of cartilage and bone formation. Curr. Opin. CellBiol. 13: 721-727.

DELISE, A.M., FISCHER, L. and TUAN, R.S. (2000). Cellular interactions andsignaling in cartilage development. Osteoarthritis. Cartilage. 8: 309-334.

FOSTER, J.W., DOMINGUEZ-STEGLICH, M.A., GUIOLI, S., KOWK, G., WELLER,P.A., STEVANOVIC, M., WEISSENBACH, J., MANSOUR, S., YOUNG, I.D.,GOODFELLOW, P.N. and SCHAFER, A.J. (1994). Campomelic dysplasia andautosomal sex reversal caused by mutations in an SRY-related gene. Nature372: 525-530.

HEGERT, C., KRAMER, J., HARGUS, G., MÜLLER, J., GUAN, K., WOBUS, A.M.,MÜLLER, P.K. and ROHWEDEL, J. (2002). Differentiation plasticity ofchondrocytes derived from mouse embryonic stem cells. J. Cell Sci. 115: 4617-4628.

HOUSTON, C.S., OPITZ, J.M., SPRANGER, J.W., MACPHERSON, R.I., REED,M.H., GILBERT, E.F., HERRMANN, J. and SCHINZEL, A. (1983). Thecampomelic syndrome: review, report of 17 cases, and follow-up on thecurrently 17-year-old boy first reported by Maroteaux et al. in 1971. Am. J. Med.Genet. 15: 3-28.

HUANG, W., CHUNG, U.I., KRONENBERG, H.M. and DE CROMBRUGGHE, B.(2001). The chondrogenic transcription factor Sox9 is a target of signaling by theparathyroid hormone-related peptide in the growth plate of endochondralbones. Proc. Natl. Acad. Sci. USA 98: 160-165.

KIST, R., SCHREWE, H., BALLING, R. and SCHERER, G. (2002). Conditionalinactivation of Sox9: a mouse model for campomelic dysplasia. Genesis. 32:121-123.

KRAMER, J., HEGERT, C., GUAN, K., WOBUS, A.M., MÜLLER, P.K. andROHWEDEL, J. (2000). Embryonic stem cell-derived chondrogenic differentia-tion in vitro: activation by BMP-2 and BMP-4. Mech. Dev. 92: 193-205.

KRAMER, J., KLINGER, M., KRUSE, C., FAZA, M., HARGUS, G. and ROHWEDEL,J. (2005). Ultrastructural analysis of mouse embryonic stem cell-derivedchondrocytes. Anat. Embryol. (Berl). 210: 175-185.

KUHN, R., RAJEWSKY, K. and MÜLLER, W. (1991). Generation and analysis ofinterleukin-4 deficient mice. Science 254: 707-710.

LANSKE, B., KARAPLIS, A.C., LEE, K., LUZ, A., VORTKAMP, A., PIRRO, A.,KARPERIEN, M., DEFIZE, L.H., HO, C., MULLIGAN, R.C., ABOU-SAMRA,A.B., JUPPNER, H., SEGRE, G.V. and KRONENBERG, H.M. (1996). PTH/PTHrP receptor in early development and Indian hedgehog-regulated bonegrowth. Science 273: 663-666.

LEFEBVRE, L., DIONNE, N., KARASKOVA, J., SQUIRE, J.A. and NAGY, A.(2001). Selection for transgene homozygosity in embryonic stem cells results inextensive loss of heterozygosity. Nat. Genet. 27: 257-258.

LEFEBVRE, V., HUANG, W., HARLEY, V.R., GOODFELLOW, P.N. and DECROMBRUGGHE, B. (1997). SOX9 is a potent activator of the chondrocyte-specific enhancer of the pro alpha1(II) collagen gene. Mol. Cell Biol. 17: 2336-2346.

LEFEBVRE, V., LI, P. and DE CROMBRUGGHE, B. (1998). A new long form ofSox5 (L-Sox5), Sox6 and Sox9 are coexpressed in chondrogenesis andcooperatively activate the type II collagen gene. EMBO J. 17: 5718-5733.

LEFEBVRE, V., ZHOU, G., MUKHOPADHYAY, K., SMITH, C.N., ZHANG, Z.,EBERSPAECHER, H., ZHOU, X., SINHA, S., MAITY, S.N. and DECROMBRUGGHE, B. (1996). An 18-base-pair sequence in the mouseproalpha1(II) collagen gene is sufficient for expression in cartilage and bindsnuclear proteins that are selectively expressed in chondrocytes. Mol. Cell Biol.16: 4512-4523.

MANSOUR, S., HALL, C.M., PEMBREY, M.E. and YOUNG, I.D. (1995). A clinicaland genetic study of campomelic dysplasia. J. Med. Genet. 32: 415-420.

MATISE, M. P., AUERBACH, W., and JOYNER, A. L. 2000. Production of TargetedEmbryonic Stem Cell Clones. In Gene Targeting: A Practical Approach (JOYNER,

A. L., ed.). Oxford University Press, Oxford, pp.101-132.

METSÄRANTA, M., TOMAN, D., DE CROMBRUGGHE, B. and VUORIO, E.(1991). Mouse type II collagen gene. Complete nucleotide sequence, exonstructure, and alternative splicing. J. Biol. Chem. 266: 16862-16869.

MOFTAH, M.Z., DOWNIE, S.A., BRONSTEIN, N.B., MEZENTSEVA, N., PU, J.,MAHER, P.A. and NEWMAN, S.A. (2002). Ectodermal FGFs induce perinodularinhibition of limb chondrogenesis in vitro and in vivo via FGF receptor 2. Dev.Biol. 249: 270-282.

MORTENSEN, R.M., CONNER, D.A., CHAO, S., GEISTERFER-LOWRANCE,A.A. and SEIDMAN, J.G. (1992). Production of homozygous mutant ES cellswith a single targeting construct. Mol. Cell Biol. 12: 2391-2395.

NG, L.J., WHEATLEY, S., MUSCAT, G.E., CONWAY-CAMPBELL, J., BOWLES,J., WRIGHT, E., BELL, D.M., TAM, P.P., CHEAH, K.S. and KOOPMAN, P.(1997). SOX9 binds DNA, activates transcription, and coexpresses with type IIcollagen during chondrogenesis in the mouse. Dev. Biol. 183: 108-121.

OBERLENDER, S.A. and TUAN, R.S. (1994). Expression and functional involve-ment of N-cadherin in embryonic limb chondrogenesis. Development 120: 177-187.

PERL, A.K., KIST, R., SHAN, Z., SCHERER, G. and WHITSETT, J.A. (2005).Normal lung development and function after Sox9 inactivation in the respiratoryepithelium. Genesis. 41: 23-32.

SEKIYA, I., TSUJI, K., KOOPMAN, P., WATANABE, H., YAMADA, Y., SHINOMIYA,K., NIFUJI, A. and NODA, M. (2000). SOX9 enhances aggrecan gene promoter/enhancer activity and is upregulated by retinoic acid in a cartilage-derived cellline, TC6. J. Biol. Chem. 275: 10738-10744.

SÜDBECK, P., SCHMITZ, M.L., BAEUERLE, P.A. and SCHERER, G. (1996). Sexreversal by loss of the C-terminal transactivation domain of human SOX9. Nat.Genet. 13: 230-232.

TANIGUCHI, M., SANBO, M., WATANABE, S., NARUSE, I., MISHINA, M. andYAGI, T. (1998). Efficient production of Cre-mediated site-directed recombi-nants through the utilization of the puromycin resistance gene, pac: a transientgene-integration marker for ES cells. Nucleic Acids Res. 26: 679-680.

WAGNER, T., WIRTH, J., MEYER, J., ZABEL, B., HELD, M., ZIMMER, J.,PASANTES, J., BRICARELLI, F.D., KEUTEL, J., HUSTERT, E., WOLF, U.,TOMMERUP, N., SCHEMPP, W. and SCHERER, G. (1994). Autosomal sexreversal and campomelic dysplasia are caused by mutations in and around theSRY-related gene SOX9. Cell 79: 1111-1120.

WEGNER, M. (1999). From head to toes: the multiple facets of Sox proteins. NucleicAcids Res. 27: 1409-1420.

WEINHOLD, B., SCHRATT, G., ARSENIAN, S., BERGER, J., KAMINO, K.,SCHWARZ, H., RUTHER, U. and NORDHEIM, A. (2000). Srf(-/-) ES cellsdisplay non-cell-autonomous impairment in mesodermal differentiation. EMBOJ. 19: 5835-5844.

WRIGHT, E., HARGRAVE, M.R., CHRISTIANSEN, J., COOPER, L., KUN, J.,EVANS, T., GANGADHARAN, U., GREENFIELD, A. and KOOPMAN, P.(1995). The Sry-related gene Sox9 is expressed during chondrogenesis inmouse embryos. Nat. Genet. 9: 15-20.

YAN, Y.L., MILLER, C.T., NISSEN, R.M., SINGER, A., LIU, D., KIRN, A., DRAPER,B., WILLOUGHBY, J., MORCOS, P.A., AMSTERDAM, A., CHUNG, B.C.,WESTERFIELD, M., HAFFTER, P., HOPKINS, N., KIMMEL, C.,POSTLETHWAIT, J.H. and NISSEN, R. (2002). A zebrafish sox9 gene requiredfor cartilage morphogenesis. Development 129: 5065-5079.

ZANETTI, N.C. and SOLURSH, M. (1986). Epithelial effects on limb chondrogen-esis involve extracellular matrix and cell shape. Dev. Biol. 113: 110-118.

ZHAO, Q., EBERSPAECHER, H., LEFEBVRE, V. and DE CROMBRUGGHE, B.(1997). Parallel expression of Sox9 and Col2a1 in cells undergoing chondro-genesis. Dev. Dyn. 209: 377-386.

ZHOU, G., LEFEBVRE, V., ZHANG, Z., EBERSPAECHER, H. and DECROMBRUGGHE, B. (1998). Three high mobility group-like sequences withina 48-base pair enhancer of the Col2a1 gene are required for cartilage-specificexpression in vivo. J. Biol. Chem. 273: 14989-14997.


Related, previously published Int. J. Dev. Biol. articles

See our Special Issue Ear Development edited by Fernando Giraldez and Bernd Fritzsch at:http://www.ijdb.ehu.es/web/contents.php?vol=51&issue=6-7

Expression of N-cadherin, N-CAM, fibronectin and tenascin is stimulated by TGF-beta1, beta2, beta3 andbeta5 during the formation of precartilage condensations.J Chimal-Monroy and L Díaz de LeónInt. J. Dev. Biol. (1999) 43: 59-67

Immunodetection of the transforming growth factors beta 1 and beta 2 in the developing murine palate.A L Gehris, M D’Angelo and R M GreeneInt. J. Dev. Biol. (1991) 35: 17-24

A change in response to Bmp signalling precedesectodermal fate choiceChris T. Dee, Abigail Gibson, Andrea Rengifo, Shun-Kuo Sun,Roger K. Patient and Paul J. ScottingInt. J. Dev. Biol. (2007) 51: 79-84

BMP2/4 and BMP5-8 in jellyfish development andtransdifferentiationSusanne Reber-Müller, Ruth Streitwolf-Engel, Nathalie Yanze,Volker Schmid, Michael Stierwald, Michael Erb and Katja SeipelInt. J. Dev. Biol. (2006) 50: 377-384

Selection and amplification of a bone marrow cell populationand its induction to the chondro-osteogenic lineage by rhOP-1: an in vitro and in vivo study.J A Andrades, J A Santamaría, M E Nimni and J BecerraInt. J. Dev. Biol. (2001) 45: 689-693

2006 ISI **Impact Factor = 3.577**

Loss of Sox9 function results in defective chondrocyte differentiation of mouse embryonic stem cells...

Documents

Transcript of Loss of Sox9 function results in defective chondrocyte differentiation of mouse embryonic stem cells...